MicroBayesAge: A Maximum Likelihood Approach to Predict Epigenetic Age Using Microarray Data

Certain epigenetic modifications, such as the methylation of CpG sites, can serve as biomarkers for chronological age. Previously, we introduced our BayesAge frameworks for accurate age prediction through the use of locally weighted scatterplot smoothing (LOWESS) to capture the non-linear relationship between methylation or gene expression and age, and Maximum Likelihood Estimation (MLE) for bulk bisulfite and RNA sequencing data. Here we now introduce MicroBayesAge, a framework that enhances prediction accuracy by subdividing input data into age-specific co-horts and employing a new two-stage process for training and testing. Age prediction for younger patients was significantly improved. MicroBayesAge also exhibited minimal bias in its age predictions. Additionally, we explored the performance of our model for sex-specific age prediction which revealed slight improvements in accuracy for male patients, while no changes were observed for female patients. MicroBayesAge provides more accurate age predictions by accounting for variations in epigenetic markers of aging among different subgroups, which have been over-looked by commonly used models.